Research

Original Investigation

Association of Football Subconcussive Head Impacts

With Ocular Near Point of Convergence
Keisuke Kawata, MS; Leah H. Rubin, PhD, MPH; Jong Hyun Lee; Thomas Sim; Masahiro Takahagi, MEd;
Victor Szwanki, MS; Al Bellamy, MS; Kurosh Darvish, PhD; Soroush Assari, BS, MS; Jeffrey D. Henderer, MD;
Ryan Tierney, PhD; Dianne Langford, PhD
Invited Commentary
IMPORTANCE An increased understanding of the relationship between subconcussive head

impacts and near point of convergence (NPC) ocular-motor function may be useful in
delineating traumatic brain injury.

Supplemental content at
jamaophthalmology.com

OBJECTIVE To investigate whether repetitive subconcussive head impacts during preseason

football practice cause changes in NPC.
DESIGN, SETTING, AND PARTICIPANTS This prospective, observational study of 29 National
Collegiate Athletic Association Division I football players included baseline and preseason
practices (1 noncontact and 4 contact), and postseason follow-up and outcome measures
were obtained for each time. An accelerometer-embedded mouthguard measured head
impact kinematics. Based on the sum of head impacts from all 5 practices, players were
categorized into lower (n = 7) or higher (n = 22) impact groups.
EXPOSURES Players participated in regular practices, and all head impacts greater than
10g from the 5 practices were recorded using the i1Biometerics Vector mouthguard
(i1 Biometrics Inc).
MAIN OUTCOMES AND MEASURES Near point of convergence measures and symptom scores.
RESULTS A total of 1193 head impacts were recorded from 5 training camp practices in the 29
collegiate football players; 22 were categorized into the higher-impact group and 7 into the
lower-impact group. There were significant differences in head impact kinematics between
lower- and higher-impact groups (number of impacts, 6 vs 41 [lower impact minus higher
impact = 35; 95% CI, 21-51; P < .001]; linear acceleration, 99g vs 1112g [lower impact minus
higher impact= 1013; 95% CI, 621 1578; P < .001]; angular acceleration, 7589 radian/s2 vs
65 016 radian/s2 [lower impact minus higher impact= 57 427; 95% CI , 31 123-80 498;
P < .001], respectively). The trajectory and cumulative burden of subconcussive impacts on
NPC differed by group (F for group linear trend1, 238 = 12.14, P < .001 and F for group
quadratic trend1, 238 = 12.97, P < .001). In the higher-impact group, there was a linear increase
in NPC over time (B for linear trend, unstandardized coefficient [SE]: 0.76 [0.12], P < .001)
that plateaued and resolved by postseason follow-up (B for quadratic trend [SE]: 0.06
[0.008], P < .001). In the lower-impact group, there was no change in NPC over time. Group
differences were first observed after the first contact practice and remained until the final
full-gear practice. No group differences were observed postseason follow-up. There were no
differences in symptom scores between groups over time.
CONCLUSIONS AND RELEVANCE Although asymptomatic, these data suggest that repetitive
subconcussive head impacts were associated with changes in NPC. The increase in NPC
highlights the vulnerability and slow recovery of the ocular-motor system following
subconcussive head impacts. Changes in NPC may become a useful clinical tool in deciphering
brain injury severity.

JAMA Ophthalmol. doi:10.1001/jamaophthalmol.2016.1085

Published online May 12, 2016.

Author Affiliations: Department of

Neuroscience, Lewis Katz School of
Medicine, Temple University,
Philadelphia, Pennsylvania (Kawata,
Lee, Langford); Department of
Kinesiology, Temple University,
Philadelphia, Pennsylvania (Kawata,
Sim, Tierney); Department of
Psychiatry, University of Illinois at
Chicago (Rubin); Department of
Athletics, Temple University,
Philadelphia, Pennsylvania (Takahagi,
Szwanki, Bellamy); Department of
Mechanical Engineering, Temple
University, Philadelphia, Pennsylvania
(Darvish, Assari); Department of
Ophthalmology, Lewis Katz School of
Medicine, Temple University,
Philadelphia, Pennsylvania
(Henderer).
Corresponding Author: Dianne
Langford, PhD, Department of
Neuroscience, Lewis Katz School of
Medicine at Temple University, 3500
N Broad St, Medical Education and
Research Bldg, Philadelphia, PA
19140 (tdl@temple.edu).

(Reprinted) E1

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archopht.jamanetwork.com/ by a Instituto Politcnico Nacional User on 05/17/2016

Research Original Investigation

Football Subconcussive Head Impact and Ocular-Motor Function

ubconcussion can be defined as a low-magnitude head

impact that does not elicit clinical signs of concussion
but potentially causes significant long-term neurological defects.1-3 Given the concern regarding concussion, understanding the effects of repetitive subconcussive impacts is critical because subconcussive impacts occur more frequently than
concussions.4-6 American football, especially at the college
level, is the sport associated with the highest incidence of concussion (0.61 per 1000 athlete exposures)7,8; moreover, college football players are reported to endure from 950 to 1353
subconcussive head impacts per season.4,9-11
While sports-related concussion is often attributed to a
single impact, prior to the concussive blow, athletes in contact sports are frequently exposed to subconcussive head
impacts.12 Establishing a threshold for concussion based on a
single impact may be inconclusive given the variation in concussion-associated head acceleration (29g-205g).5,12 Head impact kinematics in contact sports are investigated using the
Head Impact Telemetry system 13,14 and accelerometerembedded mouthguards.15,16 However, there are substantial
knowledge gaps in the relationships between subconcussive
impact kinematics and outcome measures.
The ocular-motor system orchestrates accommodation and
vergence, and their concomitant adjustments enable individuals to visualize an object at various distances and directions.17
The near point of convergence (NPC) measures the closest point
to which one can maintain convergence while focusing on an
object before diplopia occurs.18 Previously, we reported that
frontal soccer headings immediately increased (worsened) NPC
compared with baseline, with NPC impairment persisting for
24 hours after heading.19 The near point of convergence immediately after and 24 hours after heading were higher than
in control individuals.19 Thus, we hypothesized that repetitive football subconcussive impacts would worsen NPC, particularly among players with higher magnitude and greater
frequency of impacts compared with lower-impact players.

Methods
Participants
Thirty-three National Collegiate Athletic Association Division I football players at Temple University volunteered for this
study. The study was conducted during a preseason physical
examination on June 9, 2015, a series of full contact and no
contact training camp practices (August 6-21, 2015), and postseason follow-up. Inclusion criterion was being an active football team member. Exclusion criteria included a history of
head, neck, or face injury in the previous 6 months or neurological or ocular disorders. Four players exhibited abnormally high NPC scores (9.75 cm, 10.5 cm, 11 cm, and 12.5 cm)
at preseason baseline that were higher than the team mean [SD]
baseline (5.5 [2.0] cm). Because NPC scores higher than
9.5 cm to 10 cm are considered defective or convergence
insufficiency,20,21 these players were excluded from the analysis. Data from participants diagnosed as having an orthopedic injury (n = 2) and concussion (n = 1) were excluded from
further analysis. Participants refrained from substances that
E2

Key Points
Question Do repetitive subconcussive head impacts from football
alter ocular near point of convergence?
Findings This study of 29 college football players showed that
participants who sustained higher frequencies and magnitudes of
impacts had greater convergence insufficiency compared with
those with fewer and less severe impacts. The impairment in
convergence remained constant throughout the brief study
period; however, after 3 weeks of rest, convergence normalized to
baseline levels.
Meaning The slow nature of the ocular-motor systems recovery
suggests its vulnerability to repetitive subconcussive head
impacts.

could affect their nervous system (eg, stimulants), and alcohol use was prohibited. All participants gave written
informed consent, and the Temple University institutional
review board approved the study.

Study Procedures
During preseason physical examination, participants were fitted with the Vector mouthguard (i1 Biometrics Inc) that
measured the number of hits and magnitude of head acceleration. After being briefly submerged in boiling water, the mouthguard was fitted for each players bite for a secure custom
fit. Demographic information (eg, age and years of American
football experience), Sports Concussion Assessment Tool 3
symptom checklist, and ocular-motor function (NPC) were collected. During training camp practices, head impact data were
collected from 5 practices with intervals of 3 to 4 days between
measures, starting from first noncontact (pads off), first full contact (pads on), and 3 other full-contact practices (Figure 1). Postseason follow-up measurements were taken 3 weeks after the
final game of season in a subset of participants (n = 18), and no
practices occurred during this 3-week period.
Based on the sum from 5-practice impact kinematic data
collections, players were categorized into lower- and higherimpact groups for analysis. There were at least 6 impacts, 186g,
9614 radian/s2, and 15 head injury criterion differences between the highest value of the lower-impact group and the
lowest value of the higher-impact group in the number of hits,
peak linear acceleration, peak angular acceleration, and head
injury criterion, respectively (Table). Symptom checklist and
NPC data were collected 1 to 2 hours before and 1 to 2 hours
after practices.

Instruments
Head Impact Measurement
The Vector mouthguard (i1 Biometrics Inc) was used for measuring linear and rotational head kinematics during impact. The
mouthguard uses a triaxial accelerometer (ADXL377, Analog Devices) with 200g maximum per axis to detect linear acceleration. For rotational kinematics, a triaxial rotational rate gyroscope (L3GD20H, ST Microelectrics) was used. Accelerometer
and gyroscope data were low-pass filtered at 180-Hz and 40-Hz
cutoffs, respectively. When a preset threshold was triggered
(peak linear acceleration magnitude >10g, 16 pretrigger and 80

JAMA Ophthalmology Published online May 12, 2016 (Reprinted)

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archopht.jamanetwork.com/ by a Instituto Politcnico Nacional User on 05/17/2016

jamaophthalmology.com

Football Subconcussive Head Impact and Ocular-Motor Function

Figure 1. Study Flowchart

Data collection
3-4 d interval

33 Beginning of study period

Exclusion
4 Baseline near point of
convergence > 2 SD from
team average (near point
of convergence > 10 cm:
convergence insufficiency)

Pads off (first day of

training camp)
Pads on, day 1 (first
day of full gear)

Original Investigation Research

assessment was repeated twice, and mean NPC was used for
analyses. One trained tester assessed all players at all time
points, with intrarater reliability resulting in high association
between trials of all time points (r= 0.93, P < .001).
Symptom Checklist
Participants were instructed to rate the presence of any symptom at each time point using the symptom checklist, a subset
of the Sports Concussion Assessment Tool 3.23

Pads on, day 2

Statistical Analysis

Pads on, day 3

Pads on, day 4
29 End of training camp

7 Low-impact group

22 High-impact group

From 5 practices
No. of hits: 1-16
Linear, g: 40.9-478.7
Rotational, rad/s2:
1349-22 529.9

From 5 practices
No. of hits: 22-96
Linear, g: 664.6-2735.9
Rotational, rad/s2:
32 143.2-136 554.9

7 Preseason baseline

22 Preseason baseline

7, 7 Pre vs post pads off

21, 19 Pre vs post pads off

7, 7 Pre vs post pads on, day 1

21, 20 Pre vs post pads on, day 1

7, 7 Pre vs post pads on, day 2

22, 21 Pre vs post pads on, day 2

7, 7 Pre vs post pads on, day 3

22, 21 Pre vs post pads on, day 3

7, 7 Pre vs post pads on, day 4

20, 20 Pre vs post pads on, day 4

2 Postseason follow-up

16 Postseason follow-up

Pads off indicates a noncontact practice; pads on indicates a full-contact

practice.

posttrigger samples [1024 Hz sampling frequency] with a standard hit duration of 93.75 milliseconds) data were transmitted
wirelessly and stored on a secure Internet database (Figure 2).
Near Point of Convergence
Near point of convergence was assessed based on our established protocol.19,22 Participants were seated with their head
in neutral anatomical position. No spectacles were permitted; participants wore contact lenses if needed. The accommodative ruler (Bernell Inc) rested on the participants upper
lip, and an accommodative target (reduced-size Snellen chart)
was adjusted horizontally to participants eye level. The target was moved down the length of the ruler toward the eyes
at a rate of approximately 1 to 2 cm/s. Near point of convergence was taken when the tester observed eye misalignment
or when participants verbally signaled experiencing diplopia. On verbal signal, the tester stopped moving the target and
recorded the distance between the participant and object. The
jamaophthalmology.com

Group differences (lower and higher impact) in baseline characteristics were compared using independent-sample t tests
for continuous variables and 2 tests for categorical variables.
To examine group differences in the effects of subconcussive
impacts, we conducted a series of mixed-effects regression
models (MRMs) on the primary outcomes (NPC and symptoms). The MRMs were used to accommodate repeated measurements across 12 times that were correlated with different
degrees. Relative to other analytic approaches (eg, repeated
measures analysis of variance), MRM accounts for missing data,
which increases statistical power and preserves the representativeness of the results to the larger population.24
Two separate MRMs were conducted on each outcome. The
first MRM focused on group differences in the cumulative burden of subconcussive impacts across the study duration to
determine whether the trajectory/burden differed between
lower- and higher-impact groups. To this end, we conducted
random linear and polynomial trend (quadratic) modeling, a
subclass of MRM that accounts for both individual differences at baseline and over time (which may not be linear). Variables included in the model were group (lower and higher), time
(linear trend), time time (quadratic trend), and group by trend
interactions. Significant group by trend interactions indicate
that the pattern of change in an outcome measure over time
differs between individuals in the lower- and higher-impact
groups. The second MRM identified the initial time where
group differences emerged (relative to baseline), remained, and
declined across the study duration. The focus was on the rate
of change in each outcome from baseline to each time point
(eg, pads-off 1). Variables included in the analysis were group
(low and high) and dummy variables for each time (eg, padsoff 1 prepractice, pads-off 1 postpractice, pads-on 1 prepractice, pads-on 1 postpractice, etc) and all 2-way interactions
(eg, group pads-off 1 prepractice, etc). The group time interactions are the primary interest, and significance indicates that the change in an outcome from baseline to each time
point differs for individuals in the lower- and higher-impact
group. All MRMs were analyzed with SAS, version 9.4 for Windows (SAS Institute Inc) and significance was set at P < .05.

Results
Demographic and Head Impact Kinematics
A total of 1193 head impacts were recorded from 5 training camp
practices in the 29 collegiate football players using a Vector
mouthguard (Figure 2). Demographic and kinematic data for
(Reprinted) JAMA Ophthalmology Published online May 12, 2016

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archopht.jamanetwork.com/ by a Instituto Politcnico Nacional User on 05/17/2016

E3

Research Original Investigation

Football Subconcussive Head Impact and Ocular-Motor Function

Table. Group Characteristics: Demographics, Position, and Head Impact Kinematics

Group
Players
(n = 29)

Variables

Low Impact
(n = 7)

High Impact
(n = 22)

Group Difference (95% CI)

P Value

Demographics, mean (SD)

Age, y

20.6 (1.5)

20.9 (0.9)

20.5 (1.6)

0.31 (1.62 to 0.99)

.63

BMI

30.0 (4.4)

28.3 (4.9)

30.6 (4.1)

2.33 (1.51 to 6.18)

.22

No. of previous mTBI

0.61 (0.8)

0.7 (1.1)

0.55 (0.7)

0.17 (0.91 to 0.58)

.61

Years of football
experience

9.4 (5.0)

9.14 (5.5)

9.09 (4.9)

0.05 (4.56 to 4.46)

.97

8 (28)

1 (14)

7 (32)
7 (32)

NA

NA

8 (36)

NA

NA

NA

NA

Position, No. (%)

.02

Linemen (OL, DL)

Linebacker, tight end

7 (24)

Skill players (WR, DB,

RB, QB)

12 (41)

4 (57)

2 (7)

2 (29)

Special team (K)

NA

NA

NA

NA

Head impact kinematics,

median (range)a
No. of hits

29 (1 to 96)

PLA, g

940.5 (40.9 to 2735.9)

PAA, rad/s2

6.0 (1 to 16)

41.0 (22 to 96)

99.4 (40.9 to 478.7)

56 672 (1348 to 136 554) 7589 (1348 to 22 529)

HIC

897.93 (9.8 to 6282.1)

1112.3 (664.6 to 2735.9)

35.0 (20.6 to 50.9)

1012.9 (621.3 to 1578.5)

65 016 (32 143 to 136 554) 57 427 (31 123 to 80 498)

72.1 (9.8 to 426.1)

1049.1 (439 to 6282.1)

977 (273.2 to 2294.0)

<.001
<.001
<.001
.02

Abbreviations: BMI, body mass index, calculated as weight in kilograms divided by height in meters squared; DB, defensive back; DL, defensive lineman; HIC, head
injury criterion; K, kicker; mTBI, mild traumatic brain injury; NA, not applicable; OL, offensive lineman; PAA, peak angular acceleration; PLA, peak linear acceleration;
QB, quarterback; RB, running back; WR, wide receiver.
a

Based on the sum from 5-practice impact kinematic data collections (see Methods section).

Figure 2. Head Impact Kinematic Measurements Using Mouthguard and Sideline Monitoring System
A Mouthguard

Sideline monitoring system

In-mouth sensor

Y accel

Antenna

Y gyro

3-Axis accel
Battery

X accel
X gyro
3-Axis
Z accel Z gyro
Antenna

Mouthguard

A, Triaxial accel senses linear acceleration and triaxial gyro senses angular acceleration. B, Sideline antenna receives impact data by radio transmission and store in
the network server. Accel indicates accelerometer and gyro indicates gyroscope.

each group are summarized in the Table. There were no significant differences between lower and higher groups in age, body
mass index (calculated as weight in kilograms divided by height
in meters squared), number of previous concussions, and years
of football experience. Conversely, there were significant differences in head impact kinematics between lower- and higherimpact groups (number of impacts, 6 vs 41 [lower impact
minus higher impact = 35; 95% CI, 21-51; P < .001]; linear accelE4

eration, 99g vs 1112g [lower impact minus higher impact = 1013;

95% CI, 621-1578; P < .001]; and angular acceleration, 7589 vs
65 016 radian/s2 [lower impact minus higher impact = 57 427;
95% CI, 31 12380 498; P < .001], respectively) (Table).

Near Point of Convergence

The mean NPC by impact group did not differ at baseline
(F1, 238 = 0.15, P = .70) (Figure 3). Regarding group differences,

JAMA Ophthalmology Published online May 12, 2016 (Reprinted)

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archopht.jamanetwork.com/ by a Instituto Politcnico Nacional User on 05/17/2016

jamaophthalmology.com

Football Subconcussive Head Impact and Ocular-Motor Function

Original Investigation Research

Figure 3. Near Point of Convergence Between Groups Across Time Points

9

Near Point of Convergence, cm

a
7
6
5
4
3
Higher
Lower
Polynomial (higher)
Polynomial (lower)
Baseline (higher)

2
1

Pads off

on
as
ts
e
Po
s

Po
s

tp
ra
c

tic
e

tic
e
pr
ac

tp
ra
c
Po
s

Pads on 3

Pr
e

tic
e

tic
e
pr
ac

tic
e
tp
ra
c

Pads on 2

Pr
e

tic
e
Po
s

pr
ac

tic
e
tp
ra
c

Pads on 1

Pr
e

tic
e
Po
s

pr
ac
Pr
e

tic
e
Po
s

tp
ra
c

pr
ac
Pr
e

Ba

se

lin
e

tic
e

Pads on 4

Time

the trajectory and cumulative burden of subconcussive impacts on NPC differed for the lower- and higher-impact groups
(F for group linear trend1, 238 = 12.14, P < .001 and F for
group quadratic trend1, 238 = 12.97, P < .001). In the higherimpact group, there was a significant increase (worsening) in
NPC over time (B, unstandardized coefficient for linear
trend [SE], 0.76 [0.12]; P < .001) that resolved by the postseason time point (B for quadratic trend [SE], 0.06
[0.008]; P < .001). Conversely, in the lower-impact group,
there was no change in trajectory/cumulative burden of subconcussive impacts across the study duration (B for linear
trend [SE], 0.10 [0.22]; P = .64 and B for quadratic trend
[SE], 0.002 [0.01], P = .89).
In the second MRM analysis, we aimed to identify the specific point where change occurred relative to baseline and to
determine whether changes differed for those in the lower- vs
higher-impact groups. Overall, changes in NPC relative to baseline were different for those in the lower- vs higher-impact
groups (F11, 274 = 3.66, P < .001) (Figure 3). While the change
in NPC from baseline to pads-off 1 preprepractice through the
pads-on 1 prepractice did not differ by group (P >.06), the
change in NPC from baseline to pads-on 1 postpractice through
the pads-on 4 postpractice did differ by group (P <.01). Specifically, the higher-impact group demonstrated an increase
(worsening) in NPC from baseline to the pads-on 1 postpractice (B [SE], 2.70 [0.72]; P < .001) and this increase persisted
over the remaining times (P <.01) until the pads-on 4 postpractice time (B [SE], 2.12 [0.73]; P < .01). The change in NPC
from baseline to postseason follow-up did not statistically differ between the lower- and higher-impact groups (F1, 274 = 1.46,
P = .23) (Figure 3).
Given the reduced number of data points at postseason follow-up among the higher-impact group, a series of secondary
MRM analyses were conducted on the kinematics variables to
ensure that the return of NPC to baseline levels at postseason
jamaophthalmology.com

There was a group x time point

interaction vs baseline. Polynomial
indicates quadratic trend. Linear x
time and quadratic x time interaction
(P <.001) were driven by the near
point of convergence change after
pads-on 1 postpractice in the
higher-impact group (P <.001). Pads
off indicates a noncontact practice;
pads on indicates a full-contact
practice.
a
P= .06.
b
P< .01.
c
P< .001.

follow-up was not driven by the loss of follow-up data among

players possibly receiving the highest number of impacts
(higher kinematics over time). Importantly, players returning
postseason vs players not returning postseason sustained a
similar frequency and magnitude of subconcussive head impacts across the study duration (P >.38), suggesting that the
normalized NPC to baseline levels at postseason was not driven
by players receiving the highest number and magnitude of
impacts.

Symptoms
In contrast to NPC, the trajectory of symptom scores did
not differ as a function of group F for group linear
trend1, 267 = 0.87, P = .35 (eFigure in the Supplement). Additionally, changes in symptom scores from baseline to each
time were similar for all players in the lower- and higherimpact groups (F11, 274 = 0.34, P = .97).

Discussion
To our knowledge, this is the first prospective, longitudinal cohort study of collegiate football players examining the effects
of repetitive subconcussive head impacts on NPC and selfreported symptoms. The first notable finding was that subconcussive head impacts were not associated with noticeable changes in players symptom reports, regardless of
frequency and magnitude of impacts. Second, consistent with
previous studies,25,26 we found that exposure to repetitive subconcussive impacts compromised NPC function, but only
among players in the higher-impact group. Lastly, after a
3-week rest period, postseason NPC was normalized to the preseason baseline in the higher-impact group, suggesting that
ocular-motor function has the potential to reflect subclinical
brain damage and its recovery.
(Reprinted) JAMA Ophthalmology Published online May 12, 2016

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archopht.jamanetwork.com/ by a Instituto Politcnico Nacional User on 05/17/2016

E5

Research Original Investigation

Football Subconcussive Head Impact and Ocular-Motor Function

While the team mean (SD) baseline NPC (5.5 [2.0] cm) was
consistent with our previous study with healthy young adults
(5.9 [1.6] cm),22 among players in the higher-impact group, exposure to repetitive subconcussive impacts during contact
practices increased NPC scores by approximately 29% to 38%
relative to baseline levels. This is notable, given that a 3-fold
increase in NPC has been demonstrated in both soldiers with
blast-induced mild traumatic brain injury compared with non
mild traumatic brain injury control individuals25 and among
64 athletes with concussions compared with control
individuals.26,27
When interpreting these findings, it is vital to consider the
data collection days in relation to the summer camp schedule.
Pads-off 1 practice was the first day of the camp, and players had
not sustained any significant impacts for at least the past 3
months. In the helmet-only pads-off 1 practice, most players incurred none or a few subconcussive head impacts (Table). Therefore, there was no NPC change in prepractice vs postpractice or
in preseason baseline in either higher- or lower-impact groups.
After pads-off 1, we collected the data on the first full-gear practice of the camp (pads-on 1). While prepractice NPC score from
pads-on 1 remained similar to the baseline and pads-off 1 data,
postpractice NPC scores were significantly increased in the
higher-impact group, with no change in the lower impact group.
There were 3- to 4-day intervals between each pads-on data collection. We observed consistently increased NPC scores in the
higher-impact group, even at prepractice assessments, suggesting incomplete recovery of the ocular-motor system.
When considering the validity of head injury assessment
tools, it is imperative to rule out the potential contribution of
exercise and/or fatigue in relation to outcomes. The linear increase in NPC scores over time in the high impact group argues against the alternative hypothesis that fatigue and/or exercise influenced NPC. If NPC scores were impacted by fatigue
and/or exercise, we would expect to see nonlinear fluctuations in prepractice vs postpractice scores.
Our kinematic data are consistent with several studies
using the Head Impact Telemetry system. Mean numbers of
hits per player per practice were 7.0, 7.6, and 9.4 hits in our
study, Duma et al,14 and Crisco et al,13 respectively. Similarly,
mean peak linear acceleration per hit was 30.3g, 32.0g, and
28.8g in our study, Duma et al,14 and Reynolds et al,28 respectively. Our motivation for using the mouthguard sensor was
to avoid factors that may produce considerable measurement errors in a helmet-based approach such as helmet fit and
padding type.29-31 Kinematic accuracy of the instrumented
mouthguard resulted in an excellent correlation with the
matched data from an anthropomorphic testing device (crash
test dummy).15,32 Moreover, Wu et al33 used a human soccer

ARTICLE INFORMATION
Published Online: May 12, 2016.
doi:10.1001/jamaophthalmol.2016.1085.
Author Contributions: Dr Langford and Kawata
had full access to all of the data in the study and
take responsibility for the integrity of the data and
the accuracy of the data analysis.
Study concept and design: Kawata, Takahagi,

E6

heading model to test the kinematic accuracy among headgearmounted, mouthguard, and skin patch sensors, compared them
with high-speed video, and showed that mouthguard displacements were less than 1 mm, whereas headgear and skin patch
displaced as much as 13 mm and 4 mm from the ear canal reference points, respectively.
While larger prospective studies are needed to replicate our
findings, this study detected changes in NPC over time as a function of group (higher vs lower impact). The 29 players studied
contributed 325 observations for each of the primary analyses.
Moreover, our validated NPC measurement19,22 does not require administration by an experienced physician but exhibits
an excellent test-retest reliability (r = 0.93, P < .001), suggesting its potential to be used in clinical practice. Future studies
should consider additional time points for assessments, particularly during mid and late season. Additional time points
might help to determine whether NPC would continue to increase throughout the duration of the season or stabilize before returning back to baseline levels postseason. Although the
NPC measurement that we used provides a robust implication
to clinical/sideline usage, the effect of subconcussive impact on
dynamic convergence parameters, including slower peak velocity, longer latency period, and shorter duration of contractility, remain speculative.34,35 Traumatcic brain injuary produces heterogeneous signs and symptoms, and although eye
movement metrics show vulnerability in response to various
forms of traumatic brain injury,19,25,26,36 a single measure of eye
movement/alignment does not have the sensitivity and specificity to accurately assess brain damage. Therefore, multiple approaches, including blood biomarkers,37 neuroimaging,38,39 and
vestibular function,40 may be key in delineating concussion/
subconcussion pathophysiology.

Conclusions
There is growing concern that even low-level head impacts
(subconcussive) can cause significant injury if sustained repetitively. While behavioral changes in response to subconcussive head impacts are difficult to measure, evidence coupled
with neuroimaging data suggest that the ocular-motor system is particularly vulnerable and sensitive to head
impacts.19,38,39,41 Our data provide evidence of cumulative defect in the ocular NPC, and these changes may be head impact frequency and magnitude-dependent. After a 3-week rest
period, NPC normalized to the baseline. Future prospective cohort studies to investigate the clinical relevance of these NPC
changes induced by subconcussive impacts compared with
concussion are warranted.

Szwanki, Bellamy, Tierney, Langford.

Acquisition, analysis, or interpretation of data: All
authors.
Drafting of the manuscript: Kawata, Rubin, Lee, Sim,
Tierney, Langford.
Critical revision of the manuscript for important
intellectual content: Rubin, Takahagi, Szwanki,
Bellamy, Darvish, Assari, Henderer, Tierney,
Langford.

Statistical analysis: Kawata, Rubin, Lee, Sim, Assari,

Tierney,
Obtained funding: Kawata, Takahagi, Bellamy,
Langford.
Administrative, technical, or material support:
Kawata, Lee, Sim, Takahagi, Szwanki, Bellamy,
Assari, Langford.
Study supervision: Kawata, Takahagi, Szwanki,
Bellamy, Tierney, Langford.

JAMA Ophthalmology Published online May 12, 2016 (Reprinted)

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archopht.jamanetwork.com/ by a Instituto Politcnico Nacional User on 05/17/2016

jamaophthalmology.com

Football Subconcussive Head Impact and Ocular-Motor Function

Conflict of Interest Disclosures: All authors have

completed and submitted the ICMJE Form for
Disclosure of Potential Conflicts of Interest and
none were reported.
Funding/Support: Dr Kawata received support
from a research grant from the Pennsylvania
Athletic Trainers Society. Dr Langford received
support from Athole G. Jacobi, MD, the Marianne
Garman Burton Foundation for Caregivers, and a
seed grant from Temple University Office of the
Vice Provost for Research.
Role of the Funder/Sponsor: The funding sources
had no role in design or execution of the study;
collection, management, analysis, or interpretation
of the data; preparation, review, or approval of the
manuscript; or decision to submit the manuscript
for publication.
REFERENCES
1. Baugh CM, Stamm JM, Riley DO, et al. Chronic
traumatic encephalopathy: neurodegeneration
following repetitive concussive and subconcussive
brain trauma. Brain Imaging Behav. 2012;6(2):244254.
2. McKee AC, Cantu RC, Nowinski CJ, et al. Chronic
traumatic encephalopathy in athletes: progressive
tauopathy after repetitive head injury.
J Neuropathol Exp Neurol. 2009;68(7):709-735.
3. Stamm JM, Bourlas AP, Baugh CM, et al. Age of
first exposure to football and later-life cognitive
impairment in former NFL players. Neurology. 2015;
84(11):1114-1120.
4. Bailes JE, Petraglia AL, Omalu BI, Nauman E,
Talavage T. Role of subconcussion in repetitive mild
traumatic brain injury. J Neurosurg. 2013;119(5):
1235-1245.
5. Beckwith JG, Greenwald RM, Chu JJ, et al. Head
impact exposure sustained by football players on
days of diagnosed concussion. Med Sci Sports Exerc.
2013;45(4):737-746.
6. Broglio SP, Eckner JT, Martini D, Sosnoff JJ,
Kutcher JS, Randolph C. Cumulative head impact
burden in high school football. J Neurotrauma. 2011;
28(10):2069-2078.
7. Daneshvar DH, Nowinski CJ, McKee AC, Cantu
RC. The epidemiology of sport-related concussion.
Clin Sports Med. 2011;30(1):1-17, vii.
8. Gessel LM, Fields SK, Collins CL, Dick RW,
Comstock RD. Concussions among United States
high school and collegiate athletes. J Athl Train.
2007;42(4):495-503.
9. Guskiewicz KM, Mihalik JP. Biomechanics of
sport concussion: quest for the elusive injury
threshold. Exerc Sport Sci Rev. 2011;39(1):4-11.
10. Guskiewicz KM, Mihalik JP, Shankar V, et al.
Measurement of head impacts in collegiate football
players: relationship between head impact
biomechanics and acute clinical outcome after
concussion. Neurosurgery. 2007;61(6):1244-1252.
11. Schnebel B, Gwin JT, Anderson S, Gatlin R. In
vivo study of head impacts in football:
a comparison of National Collegiate Athletic

jamaophthalmology.com

Original Investigation Research

Association Division I versus high school impacts.

Neurosurgery. 2007;60(3):490-495.
12. Beckwith JG, Greenwald RM, Chu JJ, et al.
Timing of concussion diagnosis is related to head
impact exposure prior to injury. Med Sci Sports Exerc.
2013;45(4):747-754.
13. Crisco JJ, Fiore R, Beckwith JG, et al. Frequency
and location of head impact exposures in individual
collegiate football players. J Athl Train. 2010;45(6):
549-559.
14. Duma SM, Manoogian SJ, Bussone WR, et al.
Analysis of real-time head accelerations in
collegiate football players. Clin J Sport Med. 2005;15
(1):3-8.
15. Camarillo DB, Shull PB, Mattson J, Shultz R,
Garza D. An instrumented mouthguard for
measuring linear and angular head impact
kinematics in American football. Ann Biomed Eng.
2013;41(9):1939-1949.
16. King D, Hume PA, Brughelli M, Gissane C.
Instrumented mouthguard acceleration analyses
for head impacts in amateur rugby union players
over a season of matches. Am J Sports Med. 2015;
43(3):614-624.
17. Hung GK, Semmlow JL. Static behavior of
accommodation and vergence: computer
simulation of an interactive dual-feedback system.
IEEE Trans Biomed Eng. 1980;27(8):439-447.
18. Hung GK, Ciuffreda KJ, Semmlow JL. Static
vergence and accommodation: population norms
and orthoptics effects. Doc Ophthalmol. 1986;62
(2):165-179.
19. Kawata K, Tierney R, Phillips J, Jeka JJ. Effect of
repetitive sub-concussive head impacts on ocular
near point of convergence [published online
February 9, 2016]. Int J Sports Med.
20. Hamed MM, David AG, Marzieh E. The
relationship between binocular vision symptoms
and near point of convergence. Indian J Ophthalmol.
2013;61(7):325-328.
21. Von Noorden GK, Campos EC. Binocular Vision
and Ocular Motility: Theory and Management of
Strabismus. 6th ed. St. Louis, MO: Mosby; 2002.
22. Phillips J, Tierney R. Effect of target type on
near point of convergence in a healthy, active,
young adult population. J Eye Mov Res. 2015;8(3):16. doi:10.16910/jemr.8.3.1
23. McCrory P, Meeuwisse W, Aubry M, et al.
Consensus statement on Concussion in Sportthe
4th International Conference on Concussion in
Sport held in Zurich, November 2012. J Sci Med Sport.
2013;16(3):178-189.
24. Hedeker D, Gibbons RD. Longitudinal Data
Analysis. Oak Brook, IL: Wiley; 2006.
25. Cap-Aponte JE, Urosevich TG, Temme LA,
Tarbett AK, Sanghera NK. Visual dysfunctions and
symptoms during the subacute stage of
blast-induced mild traumatic brain injury. Mil Med.
2012;177(7):804-813.
26. Mucha A, Collins MW, Elbin RJ, et al. A Brief
Vestibular/Ocular Motor Screening (VOMS)
assessment to evaluate concussions: preliminary
findings. Am J Sports Med. 2014;42(10):2479-2486.

27. Pearce KL, Sufrinko A, Lau BC, Henry L, Collins

MW, Kontos AP. Near point of convergence after a
sport-related concussion: measurement reliability
and relationship to neurocognitive impairment and
symptoms. Am J Sports Med. 2015;43(12):3055-3061.
28. Reynolds BB, Patrie J, Henry EJ, et al. Practice
type effects on head impact in collegiate football.
J Neurosurg. 2016;124(2):501-510.
29. Beckwith JG, Greenwald RM, Chu JJ. Measuring
head kinematics in football: correlation between
the head impact telemetry system and Hybrid III
headform. Ann Biomed Eng. 2012;40(1):237-248.
30. Jadischke R, Viano DC, Dau N, King AI,
McCarthy J. On the accuracy of the Head Impact
Telemetry (HIT) System used in football helmets.
J Biomech. 2013;46(13):2310-2315.
31. Higgins M, Halstead PD, Snyder-Mackler L,
Barlow D. Measurement of impact acceleration:
mouthpiece accelerometer versus helmet
accelerometer. J Athl Train. 2007;42(1):5-10.
32. Bartsch A, Samorezov S, Benzel E, Miele V,
Brett D. Validation of an intelligent mouthguard
single event head impact dosimeter. Stapp Car
Crash J. 2014;58:1-27.
33. Wu LC, Nangia V, Bui K, et al. In vivo evaluation
of wearable head impact sensors. Ann Biomed Eng.
2016;44(4):1234-1245.
34. Hung GK, Ciuffreda KJ, Semmlow JL, Horng JL.
Vergence eye movements under natural viewing
conditions. Invest Ophthalmol Vis Sci. 1994;35(9):
3486-3492.
35. Hung GK, Zhu H, Ciuffreda KJ. Convergence
and divergence exhibit different response
characteristics to symmetric stimuli. Vision Res.
1997;37(9):1197-1205.
36. Ciuffreda KJ, Kapoor N, Rutner D, Suchoff IB,
Han ME, Craig S. Occurrence of oculomotor
dysfunctions in acquired brain injury:
a retrospective analysis. Optometry. 2007;78(4):
155-161.
37. Jeter CB, Hergenroeder GW, Hylin MJ, Redell
JB, Moore AN, Dash PK. Biomarkers for the
diagnosis and prognosis of mild traumatic brain
injury/concussion. J Neurotrauma. 2013;30(8):657670.
38. Johnson B, Neuberger T, Gay M, Hallett M,
Slobounov S. Effects of subconcussive head trauma
on the default mode network of the brain.
J Neurotrauma. 2014;31(23):1907-1913.
39. Johnson B, Zhang K, Hallett M, Slobounov S.
Functional neuroimaging of acute oculomotor
deficits in concussed athletes. Brain Imaging Behav.
2015;9(3):564-573.
40. Hwang S, Ma L, Kawata K, Tierney R, Jeka J.
Vestibular dysfunction following sub-concussive
head impact [published online February 17, 2016].
J Neurotrauma.
41. Johnson B, Hallett M, Slobounov S. Follow-up
evaluation of oculomotor performance with fMRI in
the subacute phase of concussion. Neurology. 2015;
85(13):1163-1166.

(Reprinted) JAMA Ophthalmology Published online May 12, 2016

Copyright 2016 American Medical Association. All rights reserved.

Downloaded From: http://archopht.jamanetwork.com/ by a Instituto Politcnico Nacional User on 05/17/2016

E7